北京邮电大学学报

  • EI核心期刊

北京邮电大学学报 ›› 2015, Vol. 38 ›› Issue (s1): 1-5.doi: 10.13190/j.jbupt.2015.s1.001

• 论文 •    下一篇

历史数据与邻居协作融合的无线传感器故障检测机制

邱雪松, 陈新颜, 杨杨, 高志鹏   

  1. 北京邮电大学 网络与交换技术国家重点实验室, 北京 100876
  • 收稿日期:2014-09-22 出版日期:2015-06-28 发布日期:2015-06-28
  • 作者简介:邱雪松(1973—), 男,教授,博士生导师,E-mail:xsqiu@bupt.edu.cn.
  • 基金资助:

    国家自然科学基金项目(61272515);教育部博士点基金项目(20110005110011)

Neighbor-Coordination in Wireless Sensor Network

QIU Xue-song, CHEN Xin-yan, YANG Yang, GAO Zhi-peng   

  1. State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications, Beijing 100876, China
  • Received:2014-09-22 Online:2015-06-28 Published:2015-06-28

摘要:

针对历史数据与邻居协作融合的思想所引生的无线传感器网络故障检测问题,提出了一种检测机制,不仅适用于故障节点随机分布的场景,也适用于故障节点集中分布的场景. 仿真结果表明:当节点度数减小、节点故障率升高时,检测精度仍可以维持在96.7%以上. 本检测机制只需要交互较少的信息,从而节省了节点的通信能耗和信道带宽.

关键词: 无线传感器网络, 故障检测, 残差比, 历史数据, 邻居协作, 集中故障

Abstract:

A new fault detection mechanism based on historical data and neighbor-coordination was presented. It gives the combination of detection methods was given based on historical data with the detection ways from neighbor-coordination. Simulation shows that the fault detection mechanism based on historical data and neighbor-coordination can be not only applied to the scene where the nodes are randomly distributed, but also for scenarios in which the failed nodes are intensive. The detection accuracy can reach 96.7% or more although the average degree of the nodes reduces or the failure rate of the nodes increases. In addition, different nodes only communicate few messages with each other in this detection mechanism. And thus, the energy of the node and the bandwidth of the channel can be saved.

Key words: wireless sensor network, fault detection, residual ratio, history data, neighbor-coordination, intensive fault

中图分类号: